With the emergence of smart LEDs, lighting based interior design is becoming popular. However, most of the smart LED-based lighting systems rely on expert-human intervention to create a desired atmosphere. For convenience, commercial lighting systems offer a number of options but their usability is fairly restricted. Therefore, an intuitive interface is required for novice users to generate the desired lighting environment. In this paper, we have developed a software, named MudGet, which automatically extracts the light mood from a digital image and controls the LED lamps to reproduce a desired lighting effect according to the extracted light mood. In our method, the light mood is regarded as a set of the representative colors of the digital image. The representative colors are extracted by utilizing K-means clustering algorithm. The dimming parameters are set for which each of the LED lamps create the lighting environment with the mood extracted by the software. To evaluate the feasibility of mood reproduction qualitatively, the degree of similarity between the light mood in the digital image and the reproduced result using LEDs is evaluated by a user study under a miniaturized experimental set. We observe that users can easily produce a desired atmosphere through the proposed MudGet software.
Highlights An image based lighting design interface is proposed. The interface controls customized LED module wirelessly. Desired lighting effect is generated from the color clustering centers of image.